Slinkie,
This is very frustrating. It is clear from your last post that you did not read any of my work. (Post # 23, which Eyesore did not quote in its entirety. Please read the original.) If you had read it, you would not think that the fact that the pellet ended up with four times the inital kinetic energy was "absurd". My analysis did not mention momentum conservation at all. It showed quantitatively that the second gun did indeed do three times more work on the pellet than the first, which negates the subsequent reasons that you had for questioning the definition of work. I will invest no more time in a thread in which other participants are unwilling to consider evidence presented to them. Debate is good, but a discussion in which my valid and quantitative points are consistently ignored or obfuscated in order to lend credence to yours, does not constitute an intelligent debate or physical analysis. Please go over my calculations again. Didn't you see how I arrived at the answer that the second gun does 375 J of work, three times more than the first?
As for your question: what is energy? It is an excellent and very deep question. I am not sure when the concept of work as defined in its precise, physical sense first arose, but consider this: there is some quantity in physics having fundamental units kgm2/s2 that as we expand physics beyond the realm of mechanics we find is conserved in every conceivable situation, although it has the capability of being transformed into various forms, some more useful for "doing stuff" than others. Examples are mechanical energy (kinetic/potential), thermal energy, EMR, etc. That is all we can say for sure. SO the definition of work as is provides for the simplest setup with which to analyse this amazing conservation law, for if you changed the definition of work to force X time, the resulting change mv caused by this new work would not be dimensionally consistent with any of the other forms of "energy" identified in other realms of physics...and hence would not qualify as energy. I see no benefit to applying the same name to totally different quantities...the accounting books would not balance. In fact, the quantities as defined in your system would not differ from their current roles in physics. They would differ in name only. What you would be calling 'work' would still have the fundamental properties of what we properly call impulse, and nothing would have changed, except that we would be left with a far more contradictory and cumbersome system of terminology in which so called "energy" would refer to at least two different types of quantities, and would therefore be neither transferable nor universally conserved.